FreshTomato Wiki

User Tools

Site Tools

Trace: • backup_script
backup_script

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
backup_script [2025/06/10 17:38] – [Goals] -Remove "Action is based on using a here doc" hogwildbackup_script [2025/09/10 15:45] (current) – Key issue fixec thilo
Line 5: Line 5:
  \\  \\
  
-  - You want to schedule a backup done via a cron job.  +  - You want to schedule a backup to be done via a cron job.  
-  - You want the script initiated from, and stored in a safe location  \\ on a server on the network, (not on the router).+  - You want the script initiated from, and stored in a safe location  \\ on a server on the network (not on the router).
   - You do not wish to install SFTP just for these backups.   - You do not wish to install SFTP just for these backups.
  
Line 13: Line 13:
 The script at the bottom of this page will create and download the backup without needing to have SFTP enabled on the router. The script at the bottom of this page will create and download the backup without needing to have SFTP enabled on the router.
  
-We could, of course, create the backup as a cron job on the router itself, and then use the mechanism applied in the script below to download the backup file. However, let's assume you want everything done in just one run of the script. To achieve this, the script covers creation of the backup in an individual file with timestamp and download.+You could, of course, create the backup as a cron job on the router itself, and then use the mechanism applied in the script below to download the backup file. However, let's assume you want everything done in just one run of the script. To achieve this, the script covers creation of the backup in an individual file with timestamp and download.
  
 In this way, just one run of the script on the backup server will create the backup and download it to to a safe location. In this way, just one run of the script on the backup server will create the backup and download it to to a safe location.
Line 41: Line 41:
  
   - The netcat command must be available on the backup server.   - The netcat command must be available on the backup server.
-  - The router's root user must have access to the public SSH key \\ of the user executing the script on backup server.+  - The user executing the script on backup server must have access to the router's root user via SSH key.
  
  \\  \\
Line 49: Line 49:
  \\  \\
  
-  - Command line arguments: backup directory, id-file and router (as IP address). \\ If provided, these will be used. Otherwise, the defaults will apply \\ (such Router IP from "ip r").+  - Command line arguments: backup directory, id-file and router (as IP address). \\ If provided, these will be used. Otherwise, the defaults will apply \\ (search Router IP from "ip r").
   - Sometimes tar may fail. You can solve this by just cleaning. The next \\ cron run may do that.   - Sometimes tar may fail. You can solve this by just cleaning. The next \\ cron run may do that.
   - All earlier backups from the same day are abandoned.   - All earlier backups from the same day are abandoned.
Line 101: Line 101:
    TRANSFER_FILENAME=config.tar    TRANSFER_FILENAME=config.tar
        
-   # It may be, that Tomato router has no - or other, wrong - timeother or no timee +   # It may happen, that Tomato router has no - or other, wrong - time 
-   # take date from localhost (i.e. backupserver) into backup filename+   # take date from localhost (i.e. backup server) into backup filename
    DATE=`date +%Y%m%d_%H%M`    DATE=`date +%Y%m%d_%H%M`
 +   
    pushd ${DIR2BACKUP} > /dev/null    pushd ${DIR2BACKUP} > /dev/null
        
Line 111: Line 111:
    # Thinks like    # Thinks like
    # VAR=`nvram get os_version`     # VAR=`nvram get os_version` 
-   # seem not to work in bash via here doc, so write results into script file and source it +   # seem not to work in bash via here doc, so write results into script file and source it. 
-   # Further the individual filename is general not known, so tar it into temp file+   # Further the individual filename is generally not known, so tar it into temp file.
    #    #
    # Kill netcat zombies    # Kill netcat zombies
    kill -9 `ps -ef | grep -v grep | grep netcat | sed -e "s/ [ ]*/ /g" | cut -d " " -f 2` 2> /dev/null    kill -9 `ps -ef | grep -v grep | grep netcat | sed -e "s/ [ ]*/ /g" | cut -d " " -f 2` 2> /dev/null
    (netcat -l -p ${PORT} > ${TRANSFER_FILENAME}) &    (netcat -l -p ${PORT} > ${TRANSFER_FILENAME}) &
 +   # Create the backup file on the router by executing the following commands (indented lines) there
    ssh ${USER}@${ROUTER} -i ${LOCAL_ID_FILE} <<ENDSSH    ssh ${USER}@${ROUTER} -i ${LOCAL_ID_FILE} <<ENDSSH
         rm -f ${SCRIPT_FILE} ${TRANSFER_FILENAME} ${PREFIX}_*_${DATE_REGEX}_${TIME_REGEX}.${EXT}         rm -f ${SCRIPT_FILE} ${TRANSFER_FILENAME} ${PREFIX}_*_${DATE_REGEX}_${TIME_REGEX}.${EXT}
Line 129: Line 130:
         tar -cvf ${TRANSFER_FILENAME} ${PREFIX}_*_${DATE_REGEX}_${TIME_REGEX}.${EXT} > /dev/null         tar -cvf ${TRANSFER_FILENAME} ${PREFIX}_*_${DATE_REGEX}_${TIME_REGEX}.${EXT} > /dev/null
         cat ${TRANSFER_FILENAME} | nc ${BACKUPHOST} ${PORT}         cat ${TRANSFER_FILENAME} | nc ${BACKUPHOST} ${PORT}
-        sleep 5 # just wait a little bit before deleting the files not needed here any more+        sleep 5 # just wait a little while before deleting the files not needed here any more
         rm -f ${SCRIPT_FILE} ${TRANSFER_FILENAME} ${PREFIX}_*_${DATE_REGEX}_${TIME_REGEX}.${EXT}         rm -f ${SCRIPT_FILE} ${TRANSFER_FILENAME} ${PREFIX}_*_${DATE_REGEX}_${TIME_REGEX}.${EXT}
    ENDSSH    ENDSSH
Line 148: Line 149:
         fi         fi
         # In total, keep only ${NO_OF_DIFF_FILES_TO_BE_KEPT} files - delete older files (of any day)         # In total, keep only ${NO_OF_DIFF_FILES_TO_BE_KEPT} files - delete older files (of any day)
-        OLDER_FILES=`ls -t ${ALL_BACKUP_FILES_PREFIX}_${DATE_REGEX}_${TIME_REGEX}.${EXT} | sed -e 1,${NO_OF_DIFF_FILES_TO_BE_KEPT}d`+        OLDER_FILES=`ls ${ALL_BACKUP_FILES_PREFIX}_${DATE_REGEX}_${TIME_REGEX}.${EXT} | sort -r | sed -e 1,${NO_OF_DIFF_FILES_TO_BE_KEPT}d`
         if [ "${OLDER_FILES}" ]; then         if [ "${OLDER_FILES}" ]; then
                 echo "keep only ${NO_OF_DIFF_FILES_TO_BE_KEPT} in total, delete:"                 echo "keep only ${NO_OF_DIFF_FILES_TO_BE_KEPT} in total, delete:"
backup_script.1749573499.txt.gz · Last modified: by hogwild

Page Tools

Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki